At any one time, approximately 6% of all Americans are in the midst of a major depressive illness. Recently, a specific syndrome called Seasonal Affective Disorder of SAd has been formally described. Specifically, SAD is characterized by recurrent fall and winter depressions with symptoms of dysthymia, fatigue, social withdrawal diminished libido, carbohydrate craving, weight gain, and hypersomnia. Bright white light can be used to alleviate the symptoms of SAD. Most commonly, phototherapy has been performed with a broad spectrum, daylight simulating, fluorescent light at 2,000 to 2,500 lux (2.3 x 10 15 photons/cm2/sec). Previous studies indicate that restricted bandwidths from the two ends of the visible spectrum (blue and red), by themselves, are unable to produce the antidepressant action of light. This implies that either the middle portion of the visible spectrum is the more active component of light in reversing SAD, or that the entire spectrum must be used simultaneously to alleviated SAD. The primary aim of this study is to determine the relative biological and therapeutic effects of tow different restricted bandwidths of visible spectrum light; green fluorescent light (501-550 nm half-peak bandwidth). Two related studies will be done with patients diagnosed with SAD. The first study is designed to determine the relative therapeutic efficacy of equal photon densities (2.3 x 10 15 photon/cm2/sec of green versus red light as assessed by the 21 item Hamilton Depression Rating Scale. Spectrum white light at an equal photon density (4.6 x 10 14 photons/cm2/sec) have equal neuroendocrine effects as measured by their capacity to suppress plasma melatonin. These studies will help determine the optimum light source for SAD phototherapy. Furthermore, these experiments will yield basic information about the photoreceptor mechanism which transmits photic stimuli to the neuroendocrine and circadian systems.